Enabling Smarter Retail and Infrastructure Automation with Infosys Video Analytics – Infosys Changes the Game with NVIDIA Metropolis
Enabling Smarter Retail and Infrastructure Automation with Infosys Video Analytics – Infosys Changes the Game with NVIDIA Metropolis
Intense competition is one of the key pressure points retailers experience today, specifically brick-and-mortar stores and big box retailers. They have been on a journey to find novel as well as more effective ways to enhance the shopping experience while improving operational efficiency.
Artificial intelligence (AI) vision analytics can help retailers strengthen their response to the growing challenges in the industry. Infosys has been helping leading retailers deliver superior shopping experience and enhanced operational efficiency through the power of video analytics.
How Infosys Video Analytics is Transforming Retail
Our solution, Infosys Video Analytics, helps retailers control shrinkage, maintain the right inventory levels and ensure store safety as well as compliance. One of our clients, a big box retailer, is using our computer vision solutions to reduce inventory loss at self-checkouts. The potential gains are estimated at over 100 million USD.
The solution helps in designing precise visual merchandising strategies and ensuring planogram compliance. Using powerful and accurate AI-powered computer vision, retailers are able to extract actionable insights from in-store analytics data. They detect the hot and cold zones, dwell times across the aisles, and conduct store traffic analysis.
We also use video analytics for augmenting the omnichannel experience, by interlinking the in-store interactions with those occurring on other channels. Leveraging the power of augmented reality and virtual reality (AR/VR) coupled with the metaverse, we enable retailers to provide a hyper-personalized shopping experience to their customers. Moreover, we build intelligent stores of the future that combine all of the above capabilities and provide a seamless, cashier-free checkout experience.
How Infosys Leverages NVIDIA Metropolis
NVIDIA Metropolis helps supercharge Infosys’ video analytics application development and deployment.
Infosys Video Analytics uses NVIDIA Metropolis software development kits (SDKs) to train and deploy AI models with GPUs. These SDKs, including NVIDIA TAO and DeepStream, and accelerators fast-track the various stages of computer vision development from building and training the models to scaling and deploying our video analytics solution.
The assortment of pre-trained models and the NVIDIA TAO low-code training framework help us reduce training efforts. Video storage toolkits and Metropolis Microservices along with DeepStream SDKs optimize the vision processing pipeline throughput and reduce overall solution cost. And with the NVIDIA Omniverse Replicator SDK, we can generate troves of synthetic data for training. This is especially important in retail, where the stock keeping unit (SKU) and packaging continuously change, because it reduces the need for frequent re-training.
Learn more about the Infosys Video Analytics solution, and about our collaboration with NVIDIA.
